African Americans die of heart disease at a rate 40 percent higher than whites. They also receive fewer invasive cardiovascular diagnostic and revascularization procedures than white patients. Addressing racial treatment disparities is of the utmost importance to the extent that they contribute to higher death rates for African Americans. While racial discrimination may play a role in determining which procedures are offered to African Americans, studies show that even when African Americans are offered the same treatment options as whites, they are less likely to agree to undergo invasive cardiovascular procedures, such as coronary artery bypass graft surgery. These findings suggest that patients' preferences may be important in explaining racial treatment disparities and warrant investigation. The immediate goal of this project is to explore racial disparities in the utilization rate of invasive cardiovascular diagnostic and revascularization procedures by examining the role that patients' risk preferences and risk perceptions play in determining their treatment choices. Risk preferences can affect medical treatment decisions, in that all medical treatments carry with them some degree of risk. Risk perceptions-individual beliefs about the risk associated with a procedure, which need not be accurate-also may influence medical treatment decisions. The specific aims of this proposal are 1) to refine and investigate the psychometric characteristics of an instrument to measure patients' belief in good and back luck; 2) to refine and investigate the characteristics of instruments to measure risk preferences in a patient population; and 3) to examine whether racial differences in risk aversion and risk perceptions explain some of the disparity between whites and African Americans in the utilization of cardiovascular surgery. The long-term goal of this proposal is to enable the candidate to become an independent health services researcher who can improve patient welfare through her study of patient preferences and how they affect medical treatment decisions. The proposed training program will build upon the candidate's background as an economist with concentrations in game theory and experimental economics, and will use the resources of the University of Pittsburgh's Graduate School of Public Health, the Center for Research on Health Care and other academic units at the University of Pittsburgh. The training will be mentored by Dr. Mark Roberts and Dr. Judith Lave.